-
Notifications
You must be signed in to change notification settings - Fork 1
/
eval.py
132 lines (118 loc) · 6.16 KB
/
eval.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
## install easse and readability packages
from easse.sari import corpus_sari
from easse.fkgl import corpus_fkgl
import numpy as np
import pandas as pd
from readability import Readability
input_file_og = open("raw_data/supreme_org_test.txt","r").read().strip().split('\n')
ref_file1_og = open("raw_data/supreme_test_labels1.txt","r").read().strip().split('\n')
ref_file2_og = open("raw_data/supreme_test_labels2.txt","r").read().strip().split('\n')
ref_file3_og = open("raw_data/supreme_test_labels3.txt","r").read().strip().split('\n')
muss_test_og = open("files/muss_test_supreme.txt","r").read().strip().split('\n')
acces_test_og = open("files/access_supreme_test.txt","r").read().strip().split('\n')[:200]
recls_outputs_og = open("files/recls_supreme_test.txt","r").read().strip().split('\n')
lsbert_outputs_og = open("files/lsbert_outputs_supreme_test.txt","r").read().strip().split('\n')
lsbert_outputs_ourcwi_og = open("files/lsbert_outputs_ourcwi_supreme_test.txt","r").read().strip().split('\n')
tst_outputs_og = open("files/gector_supreme_test.txt","r").read().strip().split("\n")
uslt_noss_og = open("files/uslt_noss_test_supreme.txt","r").read().strip().split('\n') #36.228448
uslt_ss_og = open("files/uslt_supreme_test.txt","r").read().strip().split('\n') #37.470484
scores_array = np.zeros((3,8,5))
for i in range(5):
low = i*10
high = (i+1)*10
input_file = input_file_og[low:high]
muss_test = muss_test_og[low:high]
acces_test = acces_test_og[low:high]
recls_outputs= recls_outputs_og[low:high]
lsbert_outputs = lsbert_outputs_og[low:high]
lsbert_outputs_ourcwi = lsbert_outputs_ourcwi_og[low:high]
tst_outputs = tst_outputs_og[low:high]
uslt_noss = uslt_noss_og[low:high]
uslt_ss = uslt_ss_og[low:high]
ref_file1 = ref_file1_og[low:high]
ref_file2 = ref_file2_og[low:high]
ref_file3 = ref_file3_og[low:high]
input_dc = Readability(' '.join(input_file)).dale_chall().score
muss_dc = Readability(' '.join(muss_test)).dale_chall().score
access_dc = Readability(' '.join(acces_test)).dale_chall().score
recls_dc = Readability(' '.join(recls_outputs)).dale_chall().score
lsbert_dc = Readability(' '.join(lsbert_outputs)).dale_chall().score
lsbert_ourcwi_dc = Readability(' '.join(lsbert_outputs_ourcwi)).dale_chall().score
tst_dc = Readability(' '.join(tst_outputs)).dale_chall().score
uslt_noss_dc = Readability(' '.join(uslt_noss)).dale_chall().score
uslt_dc = Readability(' '.join(uslt_ss)).dale_chall().score
muss_fkgl = corpus_fkgl(muss_test)
access_fkgl = corpus_fkgl(acces_test)
recls_fkgl = corpus_fkgl(recls_outputs)
lsbert_fkgl = corpus_fkgl(lsbert_outputs)
lsbert_ourcwi_fkgl = corpus_fkgl(lsbert_outputs_ourcwi)
tst_fkgl = corpus_fkgl(tst_outputs)
uslt_noss_fkgl = corpus_fkgl(uslt_noss)
uslt_fkgl = corpus_fkgl(uslt_ss)
# muss_fkgl = Readability(' '.join(muss_test)).flesch_kincaid().score
# access_fkgl = Readability(' '.join(acces_test)).flesch_kincaid().score
# recls_fkgl = Readability(' '.join(recls_outputs)).flesch_kincaid().score
# lsbert_fkgl = Readability(' '.join(lsbert_outputs)).flesch_kincaid().score
# lsbert_ourcwi_fkgl = Readability(' '.join(lsbert_outputs_ourcwi)).flesch_kincaid().score
# tst_fkgl = Readability(' '.join(tst_outputs)).flesch_kincaid().score
# uslt_noss_fkgl = Readability(' '.join(uslt_noss)).flesch_kincaid().score
# uslt_fkgl = Readability(' '.join(uslt_ss)).flesch_kincaid().score
muss_sari = corpus_sari(orig_sents=input_file,
sys_sents=muss_test,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
access_sari = corpus_sari(orig_sents=input_file,
sys_sents=acces_test,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
recls_sari = corpus_sari(orig_sents=input_file,
sys_sents=recls_outputs,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
lsbert_sari = corpus_sari(orig_sents=input_file,
sys_sents=lsbert_outputs,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
lsbert_ourcwi_sari = corpus_sari(orig_sents=input_file,
sys_sents=lsbert_outputs_ourcwi,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
tst_sari = corpus_sari(orig_sents=input_file,
sys_sents=tst_outputs,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
uslt_noss_sari = corpus_sari(orig_sents=input_file,
sys_sents=uslt_noss,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
uslt_sari = corpus_sari(orig_sents=input_file,
sys_sents=uslt_ss,
refs_sents=[ref_file1,
ref_file2,
ref_file3])
score_dict = {"access":[access_sari,access_fkgl,access_dc],
"muss":[muss_sari,muss_fkgl,muss_dc],
"recls":[recls_sari,recls_fkgl,recls_dc],
"lsbert":[lsbert_sari,lsbert_fkgl,lsbert_dc],
"lsbert_ourcwi":[lsbert_ourcwi_sari,lsbert_ourcwi_fkgl,lsbert_ourcwi_dc],
"tst":[tst_sari, tst_fkgl, tst_dc],
"uslt no ss":[uslt_noss_sari,uslt_noss_fkgl,uslt_noss_dc],
"uslt":[uslt_sari,uslt_fkgl,uslt_dc]}
c = 0
for key in score_dict:
for metric in range(3):
scores_array[metric,c,i] = score_dict[key][metric]
c += 1
final_score_dict = np.mean(scores_array,axis=2)
df_means = pd.DataFrame(final_score_dict,index=['SARI', 'FKGL','DC'],columns=['access','muss','recls','lsbert','lsbert_ourcwi','tst','uslt no ss','uslt'])
print(df_means)
stds = np.std(scores_array,axis=2)
df_stds = pd.DataFrame(stds,index=['SARI', 'FKGL','DC'],columns=['access','muss','recls','lsbert','lsbert_ourcwi','tst','uslt no ss','uslt'])
print(df_stds)